THE TURKISH COLONIZATION OF ANATOLIA

BY W. C. BRICE, M.A.

LECTURER IN GEOGRAPHY IN THE UNIVERSITY OF MANCHESTER

Introductory

IN the middle of the eleventh century A.D. the population of Anatolia was predominantly Christian, Greek-speaking and

sedentary. The tribes which moved into the country after the battle of Manzikert were, by contrast, Moslems of Turkish speech, who practised an economy of pastoral nomadism. What follows is a study of the nature of the Turkish colonization, with the aim of understanding how the two populations, native and immigrant, were merged. For despite the rapid establishment of Islam and the Turkish language, there is no evidence, either documentary or anthropological, of the extermination of the vanquished peoples.

Within a decade of the defeat of the main Byzantine army and the capture of the Emperor Romanus Diogenes at Manzikert in 1071, the Seljuk tribes had overrun the major part of the plateau in the face of very little resistance, and were living alongside the indigenous population. The Byzantine historians say very little 1 of how the victory was effected, but probably there was mass apostasy on the part of the Christians, and considerable intermarriage : at least, the Turkish hero Seidi Ghazi, according to legend, married a local Christian princess, possibly in order to legitimize his claim to his new territories. 2

1 See below, pp. 19 and 21-2.2 W. M. Ramsay, " The Intermixture of Races in Asia Minor, some of its

causes and effects", Proc. Brit. Acad., vii (1915-16), 3. The Behtashi hero, Haidar-es-Sultan, is also said to have married a Christian (J. W. Crowfoot, " Survivals among the Kappadolcian Kizilbash (Bektash) ", Joum. Anthrop. Inst., xxx (1900), 309). Bertrandon de la Brocquiere, who travelled in 1432-3, heard that both Ramedan, lord of Turcomania (Cilicia Campestris), and the son of the karman (prince) of Konya had Christian mothers (Travels, in W. Wright (ed.), Early Travels in Palestine (London, 1848), pp. 315, 324.)

18

TURKISH COLONIZATION OF ANATOLIA 19

The course of the Conquest

Nicetas gives a clue as to why the Anatolian population succumbed so easily; a community of Christians, he says, preferred the rule of the Turks of Konya to that of the Byzantine Emperor, and adopted Turkish customs.1 There were, in fact, profound divisions among the peoples of Anatolia at this time. In country districts the pagan cults of early Anatolia had survived, in a corrupt form, the successive attacks of Roman Emperor- worship and of orthodox Christianity, and were constantly likely to emerge in the form of heresies, such as the Montanist, Colly- ridian, and Paulician. The official disapproval and even active persecution which these provoked 2 must have antagonized whole districts and communities against the ruling power and predis­ posed them in favour of secession to a tolerant conqueror.

The ethnic and linguistic diversity of Anatolia also had the effect of weakening resistance to the invader. Although Greek was already the language of the great municipalities at the begin­ ning of the Christian era, it was slow to penetrate the rural districts, some of which retained traces of their Asianic dialects until early Byzantine times. In a moment of excitement the inhabitants of even such an accessible city as the Roman colony of Lystra were, in the first century A.D., liable to lapse from Greek into their native Lycaonian speech ;3 and until the fourth century, at least, inscriptions in the Phrygian language continued to be set up in remote parts of the western plateau. Greek speech and Hellenic civilization made particularly slow progress among the various unabsorbed ethnic stocks, on whose complex dis­ tribution in Asia Minor Strabo 4 already commented in the first century A.D., in quoting a proverb about the difficulty of fixing

1 Nicetas Choniata, Historia de lohanne Comneno, § 10 (Corpus Scriptorum Historiae Byzantinae, ed. B. G. Niebuhr, vol. 41 (Bonn, 1835), p. 50). These Christians lived on fortified islands in Lake Pusgusa (Beysehir), and in 1142 fought against John Comnenus. Nicetas concludes, OVTOI xpovcp Kparvvdev e6os yevovs Kal dpyaKelas eoriv laxvporepov.

2 Anna Comnena gives examples, Alexias, xiv, § 8-9 (Corpus Script. Hist. Byz., vol. 3 (Bonn, 1878), pp. 295-306). See also J. Laurent, " Byzance et les origines du Sultanat de Roum " in Melanges Charles Diehl, Etudes sur I'Histoire et sur I'Art de Byzance, i (Paris, 1930), 180.

3 Acts, xiv. II. 4 XII. viii. 1-4.

20 THE JOHN RYLANDS LIBRARYlimits between the Phrygians and Mysians. For example, the Gauls, according to St. Jerome,1 still spoke a Gallic language late in the fourth century, over 600 years after their arrival in Anatolia. Another instance of such a minority is afforded by the Magousaioi, against whose customs and beliefs the fourth- century Cappadocian fathers inveigh. These were probably a Persian people, transported by one of the Achaemenids to strengthen his hold on Asia Minor. Eusebius,2 who criticizes their custom of marrying within the forbidden degrees, and their pagan Persian rituals, says that in his time they were very numerous in Phrygia and Galatia.

The Empire had, after four centuries of Arab raids across the Taurus, developed an efficient strategy of frontier defence, and was able to withstand deep incursions into its territories, which at some periods were repeated almost annually. The Turkish invasion, however, presented a problem of quite different character. For while the attacks of the Arabs from their ad­ vanced bases in Cilicia and North Syria were undertaken by light raiding parties who were ready to retreat after each season's campaign,3 the Turks came to settle, and brought in the wake of their armies whole tribes, complete with families and livestock, in search of new homes and pastures.

Cahen 4 has outlined the main stages of the Seljuk conquest of Anatolia, from a study of Arab sources. According to his account, the break-through at Manzikert was simply an out­ standing episode in a long process of infiltration. Before 1071,

1 Preface to Commentary on Galatians Book H (A Select Library of Nicene and Post-Nicene Fathers, 2nd ser. vi (Oxford, 1893), 497).

2 Praeparatio Evangelica, vi (ed. F. Vigerus, Paris, 1628), 275, 279. See also St. Basil, Letters, No. 258, § 4 (A Select Library of Nicene and Post-Nicene Fathers, 2nd ser. viii (Oxford, 1895), 295-6).

3 W. M. Ramsay, " The war of Moslem and Christian for the possession of Asia Minor ", chap. VIII of Studies in the History and Art of the Eastern Provinces of the Roman Empire, ed. W. M. Ramsay (Aberdeen, 1906).

4 C. Cahen, " Les grandes lignes de 1'histoire de la penetration turque en Anatolic et en Syrie pendant la seconde moitie du XIe siecle ", in Actes da XXe Congres International des Orientalistes (Brussels, 1938), p. 336. See also J. Laurent in Melanges Diehl, op. cit. i, pp. 177-82. Laurent believes that the abandonment of the cities to the Turks in Bithynia by Nicephorus Melissenus in 1080 was the crucial surrender which enabled Soliman to set up the virtually independent Sultanate of Roum in the following year.

TURKISH COLONIZATION OF ANATOLIA 21Turcoman bands had already moved westwards from Persia, and were recruited through their chiefs as mercenary troops into both Christian and Moslem armies. In the second period, between 1071 and 1087, the organized resistance of the Byzantine Empire broke down, and small autonomous Turcoman states were set up, under local chiefs, in various parts of Anatolia and Syria. These states were weakened, however, by minor jealousies and conflicts. In the third period (1087-92), the empire of Malik Shah imposed a temporary order and unity on the new Turkish conquests from the Oxus to the Maeander; but in the fourth stage (1092-1107), this Empire broke up, and the Crusaders drove a wedge between the Turks of Anatolia and those of Syria, the former becoming united under the Seljuk state with its capital at Konya.

It seems, therefore, that the Turkish conquest did not take the form of a simple attack by a compact army, but was a more insidious penetration by pacific tribes which both preceded and followed the main fighting force. This conclusion is confirmed by Persian historians who distinguished between two classes of the Oghouz Turks when these swept on to the Iranian plateau, the first wave of wandering shepherd folk, content to seek new pastures, and the second wave of conquerors who were prepared to settle and govern. 1

Already in the early part of the twelfth century the dis­ tinction was apparent to the Byzantine historians between the disciplined Turks and the nomads, the Yiiriiks or Gocebes as they would be called in present-day Turkey. Anna Comnena 2 calls these " Turcomans", while Cinnamus 3 alludes to them as " Nomads " and " Persians ", of whom he describes about 2,000 camped in the neighbourhood of Dorylaeum. Nicetas 4 says

X W. Barthold, Turkestan down to the Mongol Invasion (" E. J. W. Gibb Memorial" Series, n.s. v), chap. II, pp. 180-322, "Central Asia down to the Twelfth Century ".

2 Anna Comnena, Alexias, xiv, § 6 (op. cit. p. 284). She says that an enemy general added to his Turkish troops reinforcements rwv Kara T-TJV *Aaiav OIKOVVTOJV TovpKOfjidvajv.

3 loannes Cinnamus, Historia, vii. § 2 (Corpus Script. Hist. Byz., vol. 26 (Bonn, 1836), p. 295).

4 Nicetas Choniata, Historia, de Manuele Comneno, iii, § 6 (op. cit. p. 163).

22 THE JOHN RYLANDS LIBRARYthey were an emigrant people who with their families moved up to the Byzantine frontiers, in search of pasture for their herds- Bertrandon de la Brocquiere l saw two types of Turcomans when he was travelling through Cilicia in the fifteenth century; the one having settled, as at Adana and in the villages near Eregli, the other retaining nomadic habits, as in the group he met near Ayas.

The process of immigration and partial settlement of nomadic tribes has continued until the present. In the last century considerable groups of Avshahrs moved into Anatolia through Armenia,2 and as late as 1951, although in consequence of strict frontier control immigration by land is no longer possible, several thousand refugee Khazak nomads came by way of sea after passing through India. Having been granted asylum in Turkey, these are faced with the same choice of settling to agriculture or retaining their nomadic economy.

The evidence of Anthropology

A recent anthropological study 3 among the Yiiriiks of southern Anatolia has shown a marked physical difference between these folk and the settled Turkish population, and, taken together with an earlier investigation by von Luschan,4 has confirmed the impression that the settled Turks are racially a mixture between the immigrant Turks and the aboriginal inhabitants of the country.

Von Luschan described in 1911 a racial variety which he recognized in a number of remote highland communities in Western Asia, ranging from the Tahtajis of Lycia to the Druses of Syria, the Yezidis of Iraq, and the Ali-Ullah-hi of Azerbaijan. This type is of short stature, with a pallid complexion, a pro­ minent nose, a tendency to hirsuteness, and a very broad and high-domed head, with the line of the neck continued vertically

1 Travels, op. cit. pp. 313-20.2 W. M. Ramsay, " The Intermixture of Races in Asia Minor . . .", op.

cit. p. 37.3 K. Giingor, Cenubt Anadolu Yuruklerinin Etno-Antropolojik Tetkiki* Ankara

(Dil ve Tarih-Cografya Fakiiltesi), 1941.4 F. von Luschan, " The Early Inhabitants of Western Asia ", Jowrn. Roy.

Anthrop. Imt., xli (1911), 221-44.

TURKISH COLONIZATION OF ANATOLIA 23into the occiput. Since this stock is found in isolated districts among peoples whose ancient beliefs have led to their ostracism by the surrounding population, it may fairly be described as the old Anatolian type. Von Luschan was wrong, however, in referring to it as aboriginal, for it is now clear that the earliest inhabitants of Anatolia were long-headed Mediterranean folk,1 and that this high-headed Armenoid type evolved during the second millennium B.C. from a mixture between these and incoming Alpine stocks. It has become exaggerated and fixed by prolonged inbreeding in isolated communities.

The Yiiriiks of southern Anatolia show physical features which are also quite uniform, and which are very similar to those of the Turks of Central Asia, themselves a very stable mixture between an old Mediterranean stock and a Mongoloid strain. The skeleton is tall and robust, the head long, the face high with prominent cheek-bones; and the eyes occasionally show a tendency towards the Mongoloid " fold ".

The settled Turkish folk,2 as might be expected, are much more variable than either of the preceding groups. Though including many individuals who closely resemble the Yuriik type, they are, on the average, shorter in stature and have more rounded faces and heads than the Yiiriiks. But the brachy- cephaly of the sedentary Turks does not often take the extreme form which is seen in the refugees of the mountain retreats ; it produces more frequently the rounded " Alpine " profile, which can be traced back in Anatolia to very early Hittite times. There is also present, especially in the towns and villages of southern Anatolia, an element of the small Mediterranean race, with short stature, delicate skeleton, dark complexion and long head, which may be aboriginal or descended from Macedonian and other colonists of Hellenistic times.

The evidence from physical anthropology is very tentative, but, such as it is, it corroborates the historical indications that the sedentary population of Turkey is usually a blend between the

1 M. §eniirek, " Anadolu Bakir ?agi ve Eti sekenesinin kraniyolojik tetkiki ", Turk Tarih Kurumu Belleten, v, 19 (1941), pp. 219-36.

2 K. Giingor quotes authorities (op. cit. p. 11, footnote [2]). See also C. S. Coon, The Races of Europe (New York, 1939), 617-22.

24 THE JOHN RYLANDS LIBRARYpre-Turkish inhabitants and the immigrant tribesfolk, with the former element predominating; while it also points to a clear ethnic difference between two groups of " outcaste " communities in Anatolia : the wandering Yiiriiks, Gocebes, and Turcomans on the one hand, little altered since their arrival from Central Asia ; and, on the other, the Bektashis, Kizilbash and Tahtajis, ancient settled folk who in isolated regions may have been little influ­ enced, physically or culturally, by the various currents of civilization which have flowed through Anatolia from Hellenistic times onwards.

The manner of the Turkish settlement

The manifest social difference between villagers and tribes- folk was accentuated by the special treatment which the latter received at the hands of the Ottoman administration. Ahmet Refik's collection of legal enactments 1 relating to the Yiiriik tribes shows that they were liable as groups to taxation and to con­ scription for special tasks, such as the building of fortifications. The scale of contributions, whether of money or labour, was based on regular censuses taken on the summer pastures, and the tribe was jointly responsible, through its headman, for fines for dereliction of duty. The administration had frequently to take stern measures to settle quarrels between adjacent tribes, and deportations on a large scale, especially to Cyprus, were not uncommon.

Settlement in Anatolia implied for the Turks not only a radical change in their material economy, but also the abandon­ ment of their tribal organization in favour of village life, which operated according to quite different customs. The tribe, which was the usual economic unit of the Central Asian Turks, com­ prised a number of clans, each of which was exogamous and recognized by its peculiar clan-crest or totem (tamga). The families, clans, tribes and larger groupings were generally arranged in multiples of ten, each with its own leader, an organization of military character which made possible rapid mustering and swift transmission of orders in times of war and crisis. The tribe had no strict territorial basis, but would generally claim

1 A. Refik, Anadoluda Turk A$iretleri (966-1200) (Istanbul, 1930).

TURKISH COLONIZATION OF ANATOLIA 25certain rights of pasturage over an ill-defined area, and any disputes on this score would be submitted to higher arbitration. The authority of the heads of clan or tribe was personal rather than official or hereditary, and election to this position was often temporary. Despite this, there was among the old Turks a feudal hierarchy, though not so much among individuals as among clans and tribes. Thus there were slave-clans as well as rich clans. 1

By contrast with the Turkish tribe, the Anatolian village was a much more self-contained community, whose limits were strictly defined, and which was organized into two strata of society. These in Phrygian times were the priests and serfs a distinction which Strabo 2 still observed at the Comanas and other old religious centres in the first century A.D. Later the high- priests were succeeded by the owners of the great fortified farms or tetrapyrgoi.3

The first stage through which the Turkish tribes passed in their adjustment to Anatolian conditions was usually one of fragmentation. This would follow from their finding much more limited stretches of territory available for pasture than in their homeland. Thus the Avshahrs,4 for example, have left one section in the neighbourhood of Lake Urmia, while another group was driven from the Uzun Yaila in the last century by a band of Circassians and took up residence in the Anti-Taurus. Quite separate from these main bodies are two villages of settled Avshahrs in south-western Phrygia in the Kara-Hiiyiik Ova. Another tribal group whose wide distribution throughout Anatolia can be traced by ethnic and place names is the Kay,5 an Oghouz tribe akin to the Ottomans, which took part in the first

1 V. V. Barthold, Les Tares d'Asie Centrale (Paris, 1945); B. Vladimirtsov, Le Regime Social des Mongols (Paris, 1948).

2 xii. ii. 3 ; XII. iii. 34; XII. iii. 37; XII. v. 3.3 W. M. Ramsay (ed. J. G. C. Anderson), The Social Basis of Roman

Power in Asia Minor (London, 1941).4 J. T. Bent, " Report to Committee on Nomad Tribes of Asia Minor ",

British Association for the Advancement of Science, Report, 1889 (Newcastle), p. 176; W. M. Ramsay, " The Intermixture of Races in Asia Minor . . .", op. cit. pp. 35, 38-9.

5 M. Fuad Kopriilu, " Kay Kabilesi hakkinda yeni notlar", Turk Tarih Kurumu Belleten, viii, 31 (1944), pp. 421-52 ; F. Demirtaf, " Osmanli devrinde Anadolu'da Kayilar ", Turk Tarih Kurumu Belleten, xii, 47 (1948), pp. 575-615.

26 THE JOHN RYLANDS LIBRARYconquest of the country, and was then dispersed to districts as far apart as Konya, Ankara and Mente§e.

The settlement of these tribal groups has taken place, at least in modern times, not by the piecemeal attachment of separate families to existing villages, as seems to happen on the northern fringe of the Syrian desert,1 but through a general decision of the whole group, under the leadership of the headman, who may be under pressure from the administration.

An example is afforded by a village of Pontus in the bend of the Halys. 2 This area has produced no buildings, monuments or inscriptions dating from the time between the Seljuk conquest and the eighteenth century. During this period the district was occupied by nomad tribes who ruled in virtual independence of the Ottoman authority. One of these tribes, the Capan 0§lu, rose to considerable power in the eighteenth century, and encouraged Greek and Armenian artisans and farmers to migrate into their territory. Gradually the nomads themselves settled down in groups alongside the newcomers, whose economy they imitated.

In the Cilician plain, tribal settlement within the last century has usually been on feudal lines.3 Former tribal chiefs, even if defeated and coerced into settlement by the government, were often given administrative titles, and granted land on which their former tribesfolk worked as tenants. Whole villages were generally occupied by members of a single tribe, though in the case of the larger settlements craftsmen and farmers were attracted from neighbouring districts and communities. In some cases the newly created landlords have left their villages to enter politics or the professions, and the tenants have acquired the freehold of their land.

If the earliest settlement of the Seljuk tribes is assumed to have taken place in a similar fashion, by the temporary dispossession of the indigenous inhabitants, the subsequent

1 S. Lloyd and W. C. Brice, " Harran ", Anatolian Studies, i (1951), 82.2 J. A. Morrison, Ali§ar : a unit of land occupance in the Kanak Su basin

of Central Anatolia (Thesis submitted to the University of Chicago, 1938. Privately reproduced.)

3 W. Eberhard, " Nomads and Farmers in Southeastern Turkey. Problems of Settlement", Oriens, vi (1953), 32-49.

TURKISH COLONIZATION OF ANATOLIA 27settlement of the tribe in groups, and the return of the former inhabitants to attach themselves to the new villages, several char­ acteristics of the Turkish colonization of the countryside become explicable. In the first place, the shift of the site of so many places, usually over quite a short distance, may be understood. This is hard to trace in the case of villages, but is noticeable, as will be shown below,1 with many towns and cities. The old situation was often unhealthy and uncomfortable, and chosen for administrative convenience, while the Turkish settlers, then as now, would be more concerned to find clear air and fresh water. Second, the general supersession of the old village name by the clan or tribal title would follow naturally from this order of events. Third, the rapid conversion of the country to Islam and Turkish speech except in the case of some remote villages of Cappadocia which remained Greek-speaking and Christian2 can be explained if the former inhabitants had to return as suppliants to the new foundations.

When did Anatolian agriculture decline?

Most of the crafts of Turkish village life, with the exception of certain techniques of animal husbandry, must have been learnt by the newcomers from the native inhabitants. Turkish tribes in Central Asia sometimes undertake casual cultivation of cereals at certain points on their annual itinerary, but most of their requirements other than the produce of their herds are obtained by trade from oases on the fringe of their territory. In particular, they would be most unlikely to have experience of the elaborate techniques of irrigation, grafting and pruning which are necessary for growing the tree-crops olive, fig, vine, apricots, poplar and the like which are so important in the Anatolian village economy.

It is certain that the standard of cultivation in Anatolia has declined between classical times and the present. Not only have special products for which certain districts, according to Strabo,3 were famous, ceased to be known the horses of

1 Pp. 39-40.2 R. M. Dawkins, Modern Greek in Asia Minor (Cambridge, 1916). 3 XI. xiv. 9; XIII. iv. II ; XII. viii. 16.

28 THE JOHN RYLANDS LIBRARYArmenia, for example, the wine of Lydia, and the fleeces of Phrygia but particular skills have been forgotten. For instance, the technique of " caprification ", whereby the fig crop is improved by laying a branch of the wild tree on the cultivated, is described by Pliny* as if it were usual, but it is now rarely practised: and St. Paul,2 in his elaborate metaphor about the grafting of the olive, speaks of one process, the grafting of the wild branch on the cultivated stock, which is not now followed.

Since large areas of Anatolia went out of cultivation, at least temporarily, at the time of the Turkish conquest, the suspicion arises that this may have been the time when the standard of cultivation fell. There is, however, good reason to think that the decline had set in long before the arrival of the Turks.

In Anatolia of Hellenistic, Achaemenid, Phrygian and possibly also earlier times, the rural economy was closely linked with the local religion.3 In the village ftome), the centre of the life of the community was the temple (/Heron) with its college of priests. The supreme deity was the Mother Goddess (Ma, Cybele), the personification of the powers of wild nature. She was the source of all earthly life, and to her it must inevitably return. The secrets of harnessing and using the energies of the Goddess belonged to her son-spouse, Attis, Lairbenos, Sabazios or Men, as he was variously called. This knowledge was, of course, essential for the practice of the arts of husbandry and agriculture, and was revealed by the God to his earthly representatives, the priests of the temple. They in turn passed on the secrets to the populace by means of the religious plays and dramas (" mys­ teries ") which were enacted for the benefit of the initiated.

These mysteries appear to have been intended to represent the divine example, which it was the duty of men to imitate. This example related not only to matters of religious significance (ritual ceremonies, purifications and the like), but also to affairs

1 Nat. Hist. xv. 21.2 Romans, xi. 17-24. See W. M. Ramsay's discussion of this passage in

" The Olive-Tree and the Wild-Olive ", chap. IX of Pauline and other Studies in Early Christian History (London, 1908).

3 W. M. Ramsay, article " Phrygians " in vol. ixof the Ena/clopaedia of Religion and Ethics, ed. J. Hastings (Edinburgh, 1917).

TURKISH COLONIZATION OF ANATOLIA 29of social life (marriage observances, for example), and even to undertakings of a purely technical kind. Indeed, it seems that economic, social and religious obligations were not clearly dis­ tinguished by the old Phrygian society. Earthly life, in all its manifold aspects, was a mirror of the ideal heavenly existence. It was the duty of humans to copy the divine example as closely as possible. But the practice did not always match the intention the mirror was often imperfect.

The formula for dealing with a sin, crime or fault (the three kinds of misdemeanour were treated similarly) is preserved in numerous inscriptions. The penitent confesses his error, says in what it consisted, recounts his penance (usually a ritual purification, fine, or rendering of service to the temple) and exhorts the reader to be warned by his example.

In practice this meant that the priestly community had a close control over the life of the village in all its aspects, including the economic. By way of apology for this system, it should be appreciated that the skills of agriculture had been steadily perfected, by generations of experience, to suit local conditions. Moreover, the temples were not only concerned with preserving the traditional rural economy, but with improving it wherever possible. Many of the larger temples specialized in one par­ ticular skill, such as bee-keeping (at Ephesus), cattle-breeding, or goat-herding. Much controlled experiment must have been undertaken, and have contributed to the high efficiency of old Anatolian farming. The larger temples, in fact, acted as agricultural research stations.

It is known that, for over a millennium after the Macedonian conquest, the old religion and customs of rural Phrygia were gradually weakened by conflict with the ideas which derived from the Greek schools of philosophy, from the official cult of the Emperors, and from Christianity. All three were first established in the municipalities, but their influence slowly radiated into the countryside.

As the strength of the old Phrygian religion was sapped, it would follow that the rural skills, whose improvement and preservation were so closely linked with it, would decline also, long before the Turkish conquest.

30 THE JOHN RYLANDS LIBRARYOn their part, the Turks contributed to the Anatolian village

economy the practice of seasonal migration with flocks in searchof fresh pastures. The custom of semi-nomadism is now sogeneral in Turkey as to be typical of Anatolian village life. Yetthere is no evidence of it in pre-Turkish times. Strabo,1 it istrue, speaks of the predatory tribes of the Pisidian hills and the.wandering Leleges with whom they were mixed. But seasonalnomadism of the sort which is practised by both Yiiriiks andvillagers in present-day Turkey depends on the pacific state ofthe countryside and on the guarantee of unmolested movementbetween pastures. Hence the constant concern of the Ottomanadministration, as Refik's compilation 2 shows, to keep theYiiriiks under control. Outlaw hill tribes like Strabo's Pisidianswould inhibit any tendency to seasonal migration up the slopes.The high town of Olba in the Taurus may have been used as acool summer retreat by the inhabitants of Sebaste and the othercoastal towns of Cilicia Tracheia. But this is quite differentfrom seasonal movement with flocks, and although the evidenceis inconclusive, since no exhaustive search has been made, no signof pre-Turkish occupation of the yaylas or high summer pasturesof the Taurus or Anti-Taurus has yet been found. Admittedlythe seasonal occupation of a few temporary shelters by a groupfrom a village would leave little trace, but if repeated year by yearnear the same water-courses, as happens now, some residue ofoccupation would be expected to accumulate.

The semi-nomadism of Turkish villages takes many different forms. Occasionally, as in the Anti-Taurus behind Maras, the whole population vacates the winter village in early summer and occupies another village among springs and orchards higher up the mountain side. More usually only part of the village population moves uphill with the flocks in summer, to live for some weeks in tents or temporary shelters. On their return they bring down hay or oak-leaves to be stored as winter fodder. Occasionally, as on the Ala Dag, agriculture is undertaken at the summer settlement.3 At Maras, the summer excursion to the

1 XIII. vii. 3. 2 A. Refik, op. cit. 3 J. Frodin, " Les Formes de la Vie Pastorale en Turquie ", Uppsala Univ.

Geografiska Amaler, Band 26 (1944), 219-72.

TURKISH COLONIZATION OF ANATOLIA 31high pastures is no longer so much an economic necessity as a privilege of the more wealthy, who take their animals for some weeks to the pastures of the summit of the mountain-range behind the town, where they camp in semi-subterranean shelters of stones, boughs and turf. At Mugla in Caria the summer station is not on a hill slope but among orchards on the outskirts of the town. The winter excursion from Harran into the Tektek Dag is also regarded as a brief holiday; 1 and Hogarth 2 observed a pathetic instance in Lycaonia of villagers living in summer tents within sight of their houses, in an attempt to preserve the custom of the summer migration.

Hogarth regarded this as the last stage in the process of adjustment from nomadic to village life. Ramsay,3 too, looked on the custom of seasonal nomadism as a transitional stage between a nomadic and an agricultural existence. Certainly it is impossible to make a firm division between nomadic and village economy in Turkey, especially as many of the Yiiriiks, in Cilicia Tracheia for example, undertake extensive cereal cultiva­ tion in their winter quarters. But the evidence has shown that the change from nomadic to settled status in Turkey is not gradual. The distinction is clear, and marked in the Cilician plain, for example, by the granting of some administrative title, such as Kaymakam, to the head of the tribe.4 Once the change is made, the community would be deeply offended to be called " Yiiriiks ". The conclusion would seem to be that the differ­ ence between nomads and villagers in Turkey is social rather than economic. The transition from the one status to the other may be influenced by a gradual evolution in the economy of the community, but the crucial step is taken when they abandon their tribal discipline and independence and submit to the ordinary administration of the country.

Strangely, although travellers have often acknowledged the contribution which the Turkish immigrants have made to

1 S. Lloyd and W. C. Brice, op. cit. p. 82.2 D. G. Hogarth, A Wandering Scholar in the Levant (London, 18%), pp.

79-81.3 W. M. Ramsay, " The Intermixture of Races in Asia Minor . . .", op. cit.

p. 33. 4 W. Eberhard, op. cit. p. 47.

32 THE JOHN RYLANDS LIBRARYthe pastoral industry of Anatolia, the usual instance which is given, that of the breeding of the Angora goat, is more likely to have been an old Anatolian technique which they learned from the native villagers.

The naturalist Tchihatcheff,1 for example, believed that the Angora goat, with its long, white, silky fleece, was introduced by the Seljuks. He based his argument on the fact that the breed is not mentioned by any of the ancient geographers ; the wools of Lycian and Cilician goats, the only kinds spoken of as coming specially from Anatolia, were coarse and cheap. But Tchi­ hatcheff was puzzled to find the Angora goat so extremely localized in a small district of Galatia and Pontus, and ascribed this to some local quality of the water or climate. Moreover, he could find no Angora goats in the Seljuk homeland in Central Asia, the nearest comparisons being the long-haired goat of Tibet and the long-haired cat of Boukhtarma.

Strabo 2 says that special breeds of sheep were kept in Anatolia, as at Gadilonitis in Pontus, where they were protected with covers made of skins, and at Laodiceia and Colossae in Phrygia, where the raven-coloured fleeces were highly prized. The secrets of selective breeding were, therefore, known in early Anatolia, but this skill does not seem to have been carefully practised, except with horses, in the Turkish homeland, where the common " Karamanli " fat-tailed sheep and black goat are the usual domesticated varieties.

Although, therefore, the first known mention of the Angora goat is by the sixteenth-century Belgian naturalist, Pierre Belon,3 who describes it in Lycaonia between Eregli and Ismil, it seems likely that it was a local breed in early Anatolia long before the Turkish invasions. There were " goat-priests" in ancient Phrygia,4 where many of the temples, as has been seen, taught and experimented in special skills of agriculture; and their particular

1 P. de Tchihatcheff, Asie Mineure, Description Physique, Statistique et Archeologique, ii (Paris, 1856), 689-725.

2 XII. iii. 13; XII. viii. 16.3 P. Belon, Les Observations de Plttsieurs Singularitez, etc., ii. (Antwerp, 1555),

96 (quoted by Tchihatcheff, op. cit. ii. 716).4 W. M. Ramsay, " Wolf-priests, Goat-priests, Ox-priests, Bee-priests",

chap. VII of Asianic Elements in Greek Civilisation (London, 1927).

TURKISH COLONIZATION OF ANATOLIA 33duty was doubtless to study the controlled breeding of goats. It may be that the Angora stock was only beginning to be evolved by classical times, and that it was not then sufficiently famous to be known to the geographers. Sir William Ramsay heard that the secrets of keeping the race pure included periodic inter-breeding with a common black goat. 1 Tchihatcheff 2 confirms this when he says that, after considerable winter losses, the animals were kept in enclosed stables and mixed with the common goat. The Angora strain became pure again at the third generation. Tchihatcheff's evidence implies that this out- breeding was intended to keep the strain strong as well as pure.

The deterioration of the landscape of Asia Minor

Since classical times there have been major changes in the landscape of Anatolia: forests have receded, fertile lowlands have been converted into malarial swamps, and harbours have become encumbered with silt. All these processes are intimately interconnected, as modern agrarian economy has shown. The reckless destruction of trees loosens the soil on hill-slopes, which is consequently eroded and washed down in such quantities as to destroy terraced fields, to interrupt the drainage of low-lying valleys, and to fill natural harbours with rapidly growing deltas.

The recession of cultivation and depopulation of the country­ side are frequently laid to the charge of the Turkish invaders, 3 who admittedly devastated wide areas along parts of the Byzantine frontier in Phrygia.4 But it is contended that the deterioration of the countryside commenced early in the Christian era, and had almost run its complete course many centuries before the arrival of the Seljuks in Anatolia.

Before searching for human causes for these topographical changes, two possible natural causes must be examined and

X W. M. Ramsay, Impressions of Turkey during Twelve Years' Wanderings (London, 1897), p. 273.

2 Tchihatcheff, op. cit. ii, 697.3 J. Laurent, for example, contends that in 1050 Byzantine Asia had " une

importance capitale dans 1'Empire. Trente ans plus tard elle n'etait plus qu'un desert ". (Byzance et les Tares Seldjoucides, in Annales de I'Est, Nancy, 1913.)

4 C. Wilson (ed.)» Murray's Handbook for Travellers in Asia Minor, Trans­ caucasia, Persia, etc. (London, 1895), Introduction, p. [47].

3

34 THE JOHN RYLANDS LIBRARYdisqualified. In the first place, climatic desiccation might account for the death of the forests without any human inter­ vention, and might thus have set the whole process in motion. The very rich variety of the Anatolian flora and its isolation on separate mountain summits 1 indicate that the country was once better watered than at present, and that since the Ice Age the rainfall has declined; though probably not at a steady rate, and with occasional increases. The evidence of the terminal moraines of Ulu Dag (Mysian Olympus) points to the same conclusion. 2 This desiccation, however, has been a long and slow process over about 20,000 years, and there is nothing to prove that it has been accelerated sufficiently in the past 2,000 years to explain the rapid retreat of the forests during that period. On the contrary, climatological evidence shows that the rainfall in Europe generally increased slightly during the early Middle Ages. 3

Second, a rise of the land relative to the sea would produce the effect of a retreating coastline, which would leave harbours stranded. But observations show that there has, on the contrary, been a slight rise in sea-level since classical times along the south and west shores of Asia Minor : the causeway at Clazomenae near Izmir and that at Myndus, the southern breakwater at Knidos, and the quay at Pompeiopolis are all a few inches below the present sea-level.

There is abundant evidence of the destruction of trees in Asia Minor in the past two millennia. No trace, for example, now remains of two woods mentioned by Strabo, the forest round the foot of Mount Argaeus,4 and the sacred grove, 80 stadia in circumference, of Daphne 5 near Antioch ; nor of the " royal forests " spoken of by Livy 6 between Mysia and Lydia, and the forests which Xenophon 7 saw near Mus. Strabo 8 also mentions a wood in the Troad which had formerly existed, but which had disappeared by his time.

1 P. de Tchihatcheff, op. cit. ii, chap. VIII.2 S. Ennc, " Glacial evidences of the climatic variations in Turkey ", Uppsala

Univ. Geografiska Annaler, Band 34 (1952), 89-98.3 E. Huntington, " The Burial of Olympia ", Geographical Journal, xxxvi

(1910), 657-86.4 xii. ii. 7. 5 xiv. ii.6. 6 1. 54,36. 7 Katabasis, I. iv. 4. 8 XIII. i. 65.

TURKISH COLONIZATION OF ANATOLIA 35There are many examples, especially near the coasts, of the

reversion of cultivated land to the wild state. For instance, the delta of the Yesil Irmak between Samsun and Unye is covered with a scrub of wild fruit-trees, including the vine, fig, apple and pear, which were once cultivated there. In western Icel, near the ruined cities of Cilicia Tracheia, the wild olive bushes which cover the ground for many miles are not part of the natural vegetation, but the remains of groves which were once tended, as is apparent from the form of their leaves. The olive orchards in the territory of Sinope, on whose importance Strabo 1 remarks, are now only found in a very narrow belt of country along the coast. Again, Theophrastus 2 mentions that the region of Pontus was noted for the special quality of its wheat; but little is now grown there.

The lowland swamps of Anatolia are not alluded to by the classical geographers, except possibly for Strabo's mention z of the unhealthy atmosphere of Caunus in Caria, which was doubtless due to its marshy situation. The malarial swamps which now render so insalubrious the environs of ancient Amisus, Ephesus and Tarsus could not have existed while these cities were prospering, nor would Corycus, Sebaste and the other towns of Cilicia Tracheia have endured the present coastal swamps of the district, which are abandoned in summer to wild pig and mosquitoes.

Although there is no direct evidence of when the forests began to be destroyed, the swamps to be formed, and lowland cultiva­ tion to be abandoned, there are better indications for deciding when the harbours were silted; and by inference from this knowledge, the other processes may be approximately dated. The canal which joined the port of Ephesus to the retreating coast must have been constructed over several centuries,4 and was finally abandoned in the sixth century when the old town was deserted for Justinian's new foundation at Ayasoluk. Miletus began to decline in commercial importance after its capture by Alexander. It continued to exist for several centuries into the

1 XII. iii. 12. 2 Hepi <Pvru>v, VIII. iv. 5. 3 XIV. ii. 3. 4 Harbour works to counteract silting were begun in the third century B.C.

(Strabo, XIV. i. 24).

36 THE JOHN RYLANDS LIBRARYChristian era, but was probably abandoned about the same time as Ephesus, as the delta of the Maeander filled the Latmic Gulf. Seleuceia Pieria, which has a land-locked harbour and channel, filled with the drifted detritus of the Orontes, fell into its ruinous condition soon after the Moslem conquest in the seventh century. The silt which obstructed the lagoon and lower navigable course of the Cydnus, and now buries the old city of Tarsus to a depth of 15-20 feet, was doubtless already accumulating rapidly when Justinian in the sixth century was obliged to construct a canal to prevent the river from over­ flowing its banks.

All this evidence suggests that the silting of the lower courses of the Anatolian rivers began shortly before the Christian era, and by the sixth or seventh century was already sufficiently advanced to render useless many harbours of the south and west coasts. It follows that the destruction of forests and conversion of fertile lowlands into swamps took place about the same time, or slightly earlier.

It is suggested that the direct cause of these changes was the large-scale commercial exploitation of the forest, mineral and agricultural resources of Anatolia, which began in Seleucid times and was continued well into the period of Roman domination.

Strabo * describes the reckless rate at which Cyprus was depleted of its forest cover at this time, to supply the metal- furnaces and ship-building yards. Settlers were officially encouraged to destroy the trees by being granted the freehold of any ground they could clear.

The forests of Anatolia also, especially those of the seaward slopes of the Taurus and Pontus ranges, were plundered for timber for ship-building. Theophrastus 2 makes special men­ tion of the districts of Mount Ida and Cilicia, and Pliny 3 of Pontus and Bithynia as sources of ships' timbers. Simultane­ ously, the smelting of iron, silver, copper and quicksilver, which were ancient crafts in the mountains of Pontus and Taurus, but which became particularly prosperous with the opening of the country to wide commerce, made steady demands on local timber.

1 XIV. vi. 5. 2 Hepi Qvrtav, IV. v. 5. s Nat. Hist., xvi. 76.

TURKISH COLONIZATION OF ANATOLIA 37It has already been argued 1 that this was the time when

some of the old skills of Anatolian farming began to decline, along with the rustic Phrygian religion on which they were closely dependent for their improvement and continuity. One symptom of this decline was an increase in the scale of farming. The village lands, which had formerly been tilled to meet only local requirements of food, were now often combined into large estates which were farmed commercially, especially for the production of livestock. The horses of Cappadocia, the fleeces of Cilicia and the wool of Pontus found a market in other parts of the Roman Empire, and to meet new demands large areas of farmland were probably converted into ranges of pasture. Sometimes native aristocratic families, from which both the religious and secular rulers were recruited, took part in this large-scale farming, as in the case of Amyntas, the last indepen­ dent ruler of Galatia, whom Strabo mentions 2 as owning above three hundred flocks of sheep. More usually, as at Antioch of Pisidia, 3 the new consolidated estates were imperial property. The official bailiff replaced the high-priest in the position of overseer, and the presiding native goddess gave way to the deified emperor.

The result of so much commercial farming, in which whole districts were exploited for a single special product, would be that so sadly familiar in recent times in countries which have embarked on large-scale agriculture: exhaustion of the soil by tillage, or destruction of the natural grass cover by excessive grazing; and thereafter, erosion of the surface strata.

The destruction of forests continued in many districts into Turkish times. In present-day Caria large pine trees are often fired and felled by the shepherds to convert forest into pasture, despite the discouragement of the government which has made this an offence. In Kurdistan, trees are stunted and frequently destroyed by the cutting of branches in autumn, for storage as

1 Pp. 27-9. above. 2 XH. vi. I.3 W. M. Ramsay, "The Tekmoreian Guest-Friends. An anti-Christian

society on the Imperial Estates at Antioch ", chap. IX of Studies in the History and Art of the Eastern Provinces of the Roman Empire, ed. W. M. Ramsay (Aberdeen, 1906), pp. 305-9.

38 THE JOHN RYLANDS LIBRARYwinter fodder. Sir Charles Fellows, 1 who travelled in 1838, described the wasteful method of collecting turpentine in Lycia, the tapped trees being burnt and abandoned to decay under the action of the weather: Theophrastus 2 mentions a similar primitive method of forest clearance about Mount Ida. Sir Mark Sykes * saw areas of fine forest destroyed in the district of Kastamonu in order to create farm land which, however poor, would, unlike the forests, yield the administration a few taxes. But the resources of Anatolia were never so systematically exploited under Turkish rule as under Roman; if only because in the earlier domination this country was a peripheral province, of importance as a frontier and as a source of raw materials, while for the Seljuks and Ottomans it was the heart of an empire, for which it supplied administrators and soldiers.

The fate of the cities at the Seljuk ConquestThe commercial economy which developed in Anatolia in

Hellenistic times contained the seeds of its own destruction, and as the fertility and resources of the countryside were exhausted, and the imperial frontiers contracted, the cities felt their pros­ perity decline. These municipalities, with their elaborate systems of connecting roads, were artificially grafted on to the country, above all in the first few decades after the Alexandrine conquest. Ethnically and linguistically they long remained distinct from the centres of native life ; their complex adminis­ tration through trade-guilds, residential tribes and age-groups, by a variety of offices, both honorary and stipendiary, was clearly distinct from the simple theocratic rule of the towns and villages of earlier Anatolia; and the philosophies and religions which they propagated never came to satisfactory terms with the old rural cults.

As their strategic and commercial functions became less important, therefore, in the days of the Eastern Empire, these cities rapidly declined. Many of the great ports, as has been seen, had already lapsed into insignificance by the seventh

1 C. Fellows, Asia Minor and Lycia (London, 1852), p. 191.2 n€pl &VT&V, ix. ii. 7.3 M. Sykes, The Caliph's Last Heritage (London, 1915), p. 387.

TURKISH COLONIZATION OF ANATOLIA 39century A.D. In default of detailed exploration, it is less easy to date the depopulation of inland cities. But Apameia decayed rapidly after the fourth century; Antioch began to fall away from its former prosperity in the sixth century; and Colossae was already deserted in the eighth century. The surviving monu­ ments of most of the ruined cities of Anatolia largely date from the early part of the Christian era.

The Turkish conquest, in destroying the last of the colonial cities in the form in which they had survived for over a millen­ nium, gave the final blow to a commercial system which was already in its last stages. These cities, which were entirely dependent for supplies on roads and aqueducts, were able to survive repeated Arab raids, because the attackers quickly withdrew and the lines of communication could be restored. But they were highly vulnerable to the interruptions, both intentional and casual, to their supply lines which were caused by the nomadic Turks, who arrived to occupy the countryside permanently. Repeated ambushes of road traffic, or the re­ moval of stones from aqueducts could sabotage the essential supplies of these cities, and they succumbed quickly to this new threat.1

In many cases the Seljuks after an interval refounded these cities, but usually at a short distance from the old site, and frequently under a new name. Thus Dorylaeum was replaced by Eskisehir, Laodiceia by Denizli, and Apameia by Dinar. Often the site of the old city is distinguished by the epithet of " Kara ", in the sense of " mysterious ", or " awesome " : the ruins of Germaniceia, for example, are known as Kara Maras, and those of Lycandos as Kara Elbistan.

1 loannes Cinnamus (Historia, vii, § 2, op. cit. p. 295), describes the desertion of Dorylaeum: dAAa Uepaai . . . -rqv re -rroXtv els e8a<f>os /Je/JA^/zewjv dvOpatiraiv eprjuov TTavraTTaaiv cireTTonjvro /cat rd rfj^e irdvra fte^pt Krai eiri XCTTTOV rrjs TrdXai tre/avoT^TO? rj<f>dvicrav i^vos. Bertrandon de la Brocquiere (Travels, op. cit. p. 313), in 1432, found no more than three hundred houses occupied in Antioch, and they almost all by Turcoman and Arab keepers of herds ; but this city had probably been devastated as much by the Crusaders as by the Turks. J. Laurent (see p. 20, footnote 4) contends that for some years Turcomans were encamped beneath the walls of the cities of Bithynia, but were allowed to occupy them in 1080.

40 THE JOHN RYLANDS LIBRARYThese shifts of site were away from the old highways of

commerce, to which water had to be brought, and into the foothills, as at Denizli and Maras, where spring water was readily available and cool breezes tempered the heat of summer. These new foundations were provincial capitals rather then centres of commerce, and much more intimately linked with their sur­ rounding countryside than the earlier cities had been. When the Seljuks re-established trade through Anatolia, the links on their highways were not the provincial cities, which usually lay aside from the main roads, but the hans placed at regular intervals of 20-25 miles. 1 These establishments are not so much an indication of the troubled state of the country 2 as proof of the changed system of commerce and communication. Rapid travel along well-kept and garrisoned highways which ran directly between municipalities was succeeded by journeying by daily stages along a few recognized caravan routes which connected distant parts of the Seljuk dominions. 3

Islam in AnatoliaA study of the religious beliefs and customs of the present

population might be expected to shed light on the character of the Turkish colonization by showing where pre-Islamic ideas have survived most tenaciously, and how far Turkish Islam has been coloured by earlier Anatolian religions.4

Unfortunately the question is complicated by the very indeterminate state of the beliefs of the Turkish tribes when they left their homes in Central Asia. They had only recently been converted to Islam, and retained many of their earlier shamanistic beliefs, while some had also come under Manichaean, Christian, and Buddhist influences. Moreover, the religious state of Anatolia in the early eleventh century was by no means uniform,

1 0. Turan, " Selcuk Kervansaraylan ", Turfy Tarih Kurumu Belleten, x, 39 (1946), PP . 471-96.

2 As W. M. Ramsay suggested in " The Peasant God ", chap. V of Luke the Physician and other Studies in the History of Religion (London, 1908), pp. 185-7.

3 A system common in many parts of Asia in the Middle Ages. Cf. W. C. Brice, " Caravan Traffic across Asia ", Antiquity, xxviii (1954), 78-84.

4 J. H. Kramers, " Islam in Asia Minor ", Analecta Orientalia, i (Leiden, 1954), 22-32.

TURKISH COLONIZATION OF ANATOLIA 41for while orthodox Christianity prevailed in the cities, heresies of various kinds were constantly arising in the countryside.

Further, Shi'ah influences were felt in Anatolia, through the contacts which the migrating Turks naturally made with the Persian population in their passage across Khorasan, and through the preaching of wandering Sufis, and the propaganda of the envoys of Shah Ismail in the sixteenth century.

Despite these complications, it is possible to distinguish between the beliefs of two groups of heterodox people in Anatolia, who have already been classified separately on economic and anthropological grounds. On the one hand, the pastoral Yiiriiks retain many of the simple shamanistic beliefs of their Central Asian ancestors ; on the other, the village-dwelling Tahtajis and Kizilbash hold very dogmatic and eclectic views, which appear to be strongly influenced by Shi'ah beliefs.

The shamanistic customs of the nomadic tribes, which Bent* noticed as far apart as the Taurus mountains and Lake Urmia, include the common practice of divination, the placation of the spirits of trees, and the reverence for the tombs of tribal ancestors, usually placed on the summer grazing-grounds.

By contrast, Bent 2 saw in the beliefs of a series of separate refugee communities in the mountains of Western Asia evidence of the survival of a formerly more extensive pagan religion. Thus the Tahtajis of Lycia, the KJzilbash of Cappadocia, the Ansairee of Cilicia, and the Ali-Ullah-hi of Lake Urmia share beliefs in Baba Nazere as the founder of their religion, in the god-head of Ali, and in the Trinity of Ali, Mohammed and Salman-el-Farsi. All drink wine at a kind of communion service, as " the image of Ali ". These similarities may be reasonably taken as evidence of contact in each case with Shi'ah ideas; though some more specific instances, like the reverence of both the Tahtajis of Lycia and the Yezidis of Iraq for the

1 J. T. Bent, " The Yourouks of Asia Minor ", Joum. Anthrop. Inst., xx (1891), 269-76; J. T. Bent, "Report to the Committee on Nomad Tribes of Asia Minor ", Brit. Assoc. for Adv. of Sd., Report, 1889 (Newcastle), p. 176.

2 J. T. Bent, " Report to the Committee for investigating Nomad Tribes of Asia Minor and North Persia", ibid. 1890 (Leeds), p. 535; J. T. Bent, " The Ansairee of Asia Minor ", Journ. Anthrop. Inst., xx (1891), pp. 225-6.

42 THE JOHN RYLANDS LIBRARYpeacock as the embodiment of evil,1 would seem to require a definite historical explanation.

More convincing evidence of the antiquity of the beliefs of the Kizilbash-Tahtaji peoples in Anatolia is the power among them of the Bektashi sect, 2 which has its two chief centres at Kir§ehir in Cappadocia and at Elmali in Lycia, in areas where the Kizilbash and Tahtajis are most numerous. Though many of the tarikas or religious orders of Anatolian Islam were inspired by Persian contacts, others, and in particular the Bektashi, may be of indigenous origin. Certainly, semi-religious trade-guilds, of the sort which are closely associated with the tarikas, were already present in pre-Turkish Anatolia, and remained open for Christian as well as Moslem membership in Ottoman times.3 The Anatolian dervish orders had much in common with the Corybantes or priests of Cybele; both danced to the music of flutes and cymbals, and practised self-mutilation. Some of the heretic Christian communities of Anatolia who inclined to ecstatic rituals may have been the link between them.

Hasluck's 4 observation has proved the wide survival of Christian elements in Anatolian Islam, and the persistence of the sanctity of certain places into Islamic from Christian and even earlier times. The common custom of respecting a saint's tomb in Turkish villages is not, however, so easily explained as a pre- Islamic survival. In some cases, as near Siiveydiye at the Orontes mouth, where the saint is called Khidr-el-Hay, he may well have a pagan ancestry. But where, as often, he carries the name of his village, he could be the eponymous ancestor of a nomadic tribe whose tomb was transferred from the summer pastures to the new village, where he is regarded as its founder. This may have happened at the village of Hassan-dede,6 at the

1 J. T. Bent, " The Yourouks of Asia Minor ", op. cit. p. 270; J. W. Crow­ foot, " A Yezidi Rite ", Man, vol. i (1901), article no. 122.

2 J. K. Birge, The Bektashi Order of Dervishes (London, 1937).3 H. A. R. Gibb and H. Bowen, Islamic Society and the West (London, 1950),

p. 283.4 F. W. Hasluck (ed. M. M. Hasluck), Christianity and Islam wider the Sultans

(Oxford, 1929).5 J. W. Crowfoot, " Survivals among the Kappadokian Kizilbash (Bektash) ",

Journ. Anthrop. Inst, XXX (1900), 309. Crowfoot would recognize here veatiges

TURKISH COLONIZATION OF ANATOLIA 43crossing of the Halys river near Ankara. Seidi Ghazi is a more definite historical hero, and the village in Phrygia was probably built round his tomb and named after him. A puzzling instance occurs at Harran, where the local saint, Sheikh Hayat- el-Harrani, who is clearly the eponymous founder of the place, bears like the village a name of very ancient origin. He was possibly adopted by the present inhabitants, who are mostly Arabs, from their pagan predecessors on the site.

ConclusionsPresent place-names may give a clue to the character of the

Turkish colonization in particular cases. It seems generally to have happened that a Turkish town which continued on the site of its predecessor retained the old name, as at Konya, Tarsus, and Kayseri; while if, as frequently, it was refounded a short distance from the ruined place, it was re-named, for instance at Aydin (the former Tralles), Dinar (Apameia), and Mara§ (Germaniceia). Villages rarely kept their old name; generally they received a purely descriptive title, such as Akpinar or Qaylarbasi, or else they were called after the tribe which settled there, Karakeci, for example, or Ahmetli. It was probably only in a few special cases, where the village remained largely Greek- speaking, that old names survived in a mispronounced form, like Efsus (Ephesus), Tefenni (Stephanos) and Ayasoluk 1 (Hagios Theologos).

The nineteenth-century traveller W. M. Leake 2 remarked on passing a Turkish mollah who was journeying in excessive comfort " Graecia capta ferum victorem cepit ". But the taste for Byzantine luxuries only affected a very small proportion of the Turkish conquerors. The immediate result of their arrival was the final collapse of the Graeco-Roman municipal foundations on whose lavish expenditure of the resources of the country the wealth of Constantinople had long depended. The

of hero cults of the ancient Aegean world, and, more specifically, ethnic survivals of the Magousaioi (see p. 20 and footnote 2).

1 Occasionally the old pre-Greek name, which must have always continued in popular use, was revived in Turkish times, as at Edessa, the pre-Hellenistic Orhai and Turkish Urfa.

2 W. M. Leake, Journal of a Tour in Asia Minor (London, 1824), pp. 3-4.

44 THE JOHN RYLANDS LIBRARYTurks preferred village to city life, and even when they refounded the cities, on a smaller scale and on a quite different plan, most of the traders or craftsmen belonged to the millets or minorities. Village life was enriched by the Turkish arts of animal husbandry, especially those concerned with seasonal migration ; and although many of the skills which had kept the balance of rural economy in pre-Hellenistic times were never recovered, the decline in the fertility of the countryside was generally arrested after the municipalities collapsed.